Building a representative research sample for your business is more important than you probably think.
If you are like many companies, you are conducting market research among those you hope reflect your customer base, or those you wish will one day become customers. Either way, sample bias has the potential to drastically undercut the value of your data collection methods.
Assuming you understand the importance of a representative research sample, here are three common ways to achieve one.
Mostly reserved for companies with decent sized budgets, random sampling simply means you survey enough respondents from your target population at random. The key is to get a large enough sample size so that respondents from all different demographics and other defining characteristics are included, thereby minimizing potential sample bias. It’s important to note that this method can be time consuming and expensive, though it is considered a superior methodology.
Considerably more common, purposive sampling involves carefully selecting respondents for a survey based on what you know about the target population. Quotas are set for approximately three or four defining characteristics, sometimes demographic in nature. Anything more than three or four can make the research process complicated and prone to error, and is therefore not recommended. To determine what those defining characteristics are, it’s wise to examine previous research including census data and industry-specific market research reports. Purposive sampling ensures that the appropriate number of people from each characteristic are surveyed. For example, if you are a restaurant owner that knows 60 percent of your target population dine out twice a week, you might consider a quota specifying that 60 percent of survey respondents dine out twice a week.
Consistent survey timing
When, and how often, you survey your target population contributes to how representative of a sample you get. If you conduct ongoing research for brand awareness or employee satisfaction, for example, it is integral that you gather responses at consistent time intervals (e.g. hourly, weekly, monthly), on the same day (e.g. Monday, Tuesday, Wednesday), and at the same time (e.g. 8 am, 12 pm, 4 pm). All of these factors play a significant role in the accuracy of your sample and how actionable your data is.
Remember, following these tips to get a representative research sample is well worth the effort. Crafting an amazing research survey which follows all the best-practices is useless if you have an unrepresentative sample.